Comment on "Statistical Modeling: The Two Cultures" by Leo Breiman

TL;DR

This paper discusses the influence of different modeling approaches on causal inference, highlighting issues related to model complexity, estimation methods, and cultural divisions in the field.

Contribution

It provides a critical commentary on the impact of modeling choices in causal inference, expanding on Breiman's two cultures concept.

Findings

Model complexity affects causal interpretation

Plug-in vs targeted estimation trade-offs

Tuning flexible causal estimators is challenging

Abstract

Motivated by Breiman's rousing 2001 paper on the "two cultures" in statistics, we consider the role that different modeling approaches play in causal inference. We discuss the relationship between model complexity and causal (mis)interpretation, the relative merits of plug-in versus targeted estimation, issues that arise in tuning flexible estimators of causal effects, and some outstanding cultural divisions in causal inference.